JP2003044072A - Voice reading setting device, voice reading device, voice reading setting method, voice reading setting program and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To comfortably perform reading by appropriately changing the reading characteristics corresponding to the liking of a user even when special data for voice reading setting are not integrated in the data of a voice reading object. SOLUTION: A category, to which a reading object document belongs, is discriminated (step S2), voice reading setting corresponding to the category of the discriminated result is applied to the reading object document (step S3), and voice reading is performed on the basis of reading object document data and voice reading setting corresponding to the reading object document (steps S4-S6).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice reading setting device, a voice reading device, a voice reading setting method, a voice reading setting program and a recording medium.

[0002]

2. Description of the Related Art Conventionally, there has been known a voice reading device which performs voice synthesis on text data and reads it by voice. In such a voice reading device, in order to change the characteristic of the generated voice, in Japanese Patent Laid-Open No. 3-276193, the control data for changing the characteristic (such as a tag) is previously incorporated in the text data, and the text data is stored. There is disclosed a technique for setting and changing voice synthesis based on the control data when reading the. Regarding the reading of electronic mail, Japanese Patent Laid-Open No. 4-10
Japanese Patent No. 734 discloses a technique for setting voice synthesis according to the sender of an electronic mail. Further, Japanese Patent Laid-Open No. 9-265380 discloses a technique for changing the output voice according to the attribute of each text string in a document.

[0003]

By the way, the text data for automatic reading includes a tag for specifying the reading characteristic in advance, and the reading characteristic is generally set based on this tag or the like. . Examples of the read-aloud characteristics to be set include the type of voice such as a male voice, a female voice, a voice of a child, and a voice of a robot, the degree of intonation, pitch high / low, and speed. In addition, by incorporating control data for changing the type of voice into the data for setting the reading characteristics, it is possible to add a realistic atmosphere to the reading. However, in reality, it is rare that the text data includes data for setting the reading characteristics, even though the browser or the like has the automatic reading function. Is rarely done automatically.

That is, unless the user changes the setting of the reading characteristic, the voice synthesis is performed with the specified setting and the reading process is performed. Such constant reading with a constant voice increases the user's impression that the sound is a machine-synthesized sound, and sometimes causes the user to feel uncomfortable.
Therefore, an object of the present invention is to read aloud comfortably by appropriately changing the reading characteristics according to the user's preference even when the data to be read aloud does not include special voice reading setting data. It is to provide a voice reading setting device, a voice reading device, a voice reading setting method, a voice reading setting program, and a recording medium capable of performing the above.

[0005]

In order to solve the above-mentioned problems, a voice-reading setting device for performing settings for voice-reading a read-aloud document includes a discriminating section for discriminating a category to which the read-aloud document belongs, It is characterized in that it comprises a setting unit for making a voice reading setting corresponding to the category of the discrimination result for the reading target document. According to the above configuration, the determining unit determines the category to which the reading target document belongs. The setting unit performs voice reading setting corresponding to the category of the discrimination result for the reading target document.

In this case, a setting storage unit may be provided for storing in advance the voice reading setting corresponding to each category, and the setting unit may perform the voice reading setting by referring to the setting storage unit. . Further, the determination unit may include an analysis unit that analyzes the content of the reading target document. Further, when the reading target document or a part of the specific document is set as the reading target document, the analyzing unit further includes a keyword included in the specific document, a title, link destination information of the reading target document, and U.
The content of the reading target document may be analyzed based on at least one of RL and image content.

Further, the voice reading device comprises the voice reading setting device, and a voice reading unit for reading voice based on the reading target document data corresponding to the reading target document and the voice reading setting. There is. According to the above configuration, the voice reading unit performs voice reading based on the read target document data corresponding to the read target document and the voice reading setting set by the voice reading setting device.

In this case, a setting display unit for displaying the voice reading setting at the time of the voice reading, a setting changing unit for changing the voice reading setting by the user, and a change of the voice reading setting for the voice setting device. It may be configured to include a setting update unit that is reflected on the side.

Further, the voice reading setting method for setting the voice reading of the reading target document includes a determination process of determining the category to which the reading target document belongs, and a determination result category for the reading target document. It is characterized by having a setting process of performing corresponding voice reading setting. According to the above configuration, in the determination process, the category to which the reading target document belongs is determined. Then, in the setting process, voice reading setting corresponding to the category of the discrimination result is performed on the reading target document.

In this case, there is provided a setting storing process for storing in advance the voice reading setting corresponding to each category,
In the setting process, the voice reading setting may be performed with reference to the voice reading setting stored in the setting storing process. Further, the determining step may include an analyzing step of analyzing the content of the reading target document. Further, in the analysis process, when the reading target document or a part of the specific document is set as the reading target document, the keyword, the title, the link destination information of the reading target document, the URL, and the image content are included in the specific document. The content of the reading target document may be analyzed based on at least one of the above.

Further, a voice-reading setting program for setting a computer to read aloud the read-aloud document is made to discriminate the category to which the read-aloud document belongs, and the category of the discrimination result is set for the read-aloud document. Make the corresponding voice reading settings,
It is characterized by that. In this case, the voice reading setting corresponding to each category may be stored in advance, and in the setting process, the voice reading setting stored in the setting storing process may be referred to to perform the voice reading setting. Further, these voice reading setting programs may be recorded in a recording medium.

[0012]

BEST MODE FOR CARRYING OUT THE INVENTION Next, preferred embodiments of the present invention will be described with reference to the drawings. [1] First Embodiment [1.1] Configuration of First Embodiment FIG. 1 shows a schematic configuration block diagram of a document reading device according to the embodiment. The document reading device 10 has a configuration similar to that of a general computer system.
The document reading device main body 11, a data input unit 12, a display unit 13, and an output unit 14 are provided. FIG. 2 shows a schematic block diagram of the document reading device body. The document reading device main body 11 functions as a document reading setting device and a document reading device.
1, a ROM 22, a RAM 23, an external storage device 24, and a communication interface 25. M
The PU 21 uses the storage area of the RAM 23 as a work area, and based on a control program stored in the ROM 22 and a document reading application program (functions as a voice reading characteristic setting program) stored in the external storage device 24, the document reading apparatus main body 11 To control. The ROM 22 stores various non-volatile data such as control programs. The RAM 23 is used as a working area of the MPU 21, and temporarily stores an application program and various data.

The external storage device 24 is composed of a hard disk, etc., and various databases such as a category definition database, a reading setting definition database, a morphological analysis dictionary, a phoneme dictionary, a waveform dictionary, etc., which will be described later, a document reading application program, a browser program, It stores various application programs such as electronic mail programs. The communication interface 25 is L
AN, analog telephone line network, ISDN line,
It provides an interface function with various networks such as digital telephone line networks such as ADSL and optical fiber line networks. The data input section 12 is
It is equipped with a keyboard, mouse, tablet, etc., and inputs various data including document data. Display unit 13
Displays various data under the control of the document reading device body 11. The output unit 14 has a function as a voice output device that actually outputs voice based on voice data output from the document reading device body 11, and performs various output processes.

[1.2] Display Screen First, an example of the browser display screen will be described prior to the description of the specific operation. As shown in FIG. 3, the browser display screen displayed on the display unit 13 roughly includes a browser operation unit 31, a URL display unit 32, and a text display unit 33. The browser operation unit 31 includes a “return” button 41 for returning the displayed web page from the currently displayed Nth page to the (N−1) th page, and the displayed web page from the currently displayed Nth page to the (N) th page. N +
1) "Forward" button 42 for advancing to the page, "Cancel" button 43 for stopping downloading or updating of display data
In addition to various display operation buttons 44 such as the above, a reading button 45 relating to the present invention is provided. URL display section 3
In 2, the URL of the currently displayed web page is displayed. The body display section 33 is an area for displaying the body of the web page to be displayed, and in the case of FIG. 3, three frames are displayed. Specifically, in the case of FIG. 3, the first
The title "XX newspaper" is displayed in the frame, the today's news article is displayed in the second frame, and the third
The weather information for tomorrow is displayed in the frame. By the way, in the present embodiment, the reading sentence is classified into categories, and the reading is performed according to the reading setting corresponding to the category to which the reading sentence belongs.

[1.3] Category Definition Database and Reading Setting Definition Database Next, the category definition database CDDB and the reading setting definition database RDDB used for category classification and reading setting of reading sentences will be described.
FIG. 4 shows an example of the category definition database. The category definition database CDDB roughly includes category name data CN, keyword data KW, and reading setting data DN. The category name data CN is provided for the user to easily identify the corresponding category. For example, names such as “news” for the news category, “weather forecast” for the weather forecast category, “fortune-telling” for the fortune-telling category, etc. are assigned names that the user can easily understand. The name can be arbitrarily set by the user. The keyword data KW stores a plurality of keywords that are considered to appear frequently in each category. For example, in the news category, "news",
The “case”, “accident”, “trial”, etc. are stored.
In the fortune-telling category, “fortune-telling”, “fortune”, “lucky”, “Aries”, etc. are stored. Furthermore, the bulletin board (electronic bulletin board) categories include "bulletin board" and "BB".
"S", "post", etc. are stored.

The reading setting data DN stores the setting name on the reading setting definition database corresponding to the reading setting of the category. For example, in the case of the weather forecast category, the setting name = “weather caster”, and in the case of the sports category, the setting name =
Has become an "announcer". The definition of the actual reading setting corresponding to each setting name will be described later.

FIG. 5 shows an example of the reading setting definition database. The reading setting definition database RDDB is
Broadly classified, setting name data DNN and voice type data D
V, speed data DS, intonation data DI, tone data DM
Is equipped with. The setting name data DNN is provided so that the user can easily identify the reading setting, and also the category definition database CDDB.
Used for reference from. The voice type data DV stores a name assigned to specify the voice data to be used. Specifically, in the case of a female voice, "female voice A"(girl's voice), "female voice B" (younger woman's higher voice), "female voice C" (younger woman's lower voice), etc. The names assigned to the voice types are stored as shown in. The speed data DS is data for setting the speed at the time of reading, and for example, “slow”, “slightly slow”, “normal”, “slightly fast”, “fast”, etc. are stored.

The inflection data DI is data for setting the tone at the time of reading, and for example, "large", "slightly large", "normal", "slightly small", "small", etc. are stored. . The tone data DM is data for setting what tone to read aloud,
For example, "very polite", "polite", "slightly polite",
"Ordinary", "Slightly", "Slightly",
It stores "I'm very tired". In this case, the speed data DS, the intonation data DI, etc. are used to specify the duration, pitch, power, etc. of each phoneme. Further, the tone data DM is used to specify the intonation and the pronunciation method. Although not shown, examples of data that can be used in the same manner include regional data (specifying dialects, accents, etc.) and the like.

[1.4] Operation of the First Embodiment Next, the operation will be described with reference to the processing flowchart of the reading processing of FIG. In the following description, the case where a part of a sentence is read out is described as an example, but the same is true when the whole sentence is read out. First, the user displays a document including a sentence to be read aloud on the display screen of the display unit 13. In this case, the document is stored in the external storage device 24 in advance, is directly input using the data input unit 12, or is downloaded via the communication interface 25. When selecting the reading sentence, the user specifies the reading range of the sentence displayed on the display screen of the display unit 13, and clicks and depresses the reading button 45 (step S1). As a result, the MPU 21 makes the category definition database CDDB in the external storage device 24.
To perform sentence classification processing (step S2). Specifically, the MPU 21 analyzes a sentence whose reading is specified by morphological analysis using the morphological analysis dictionary DIC1 and extracts a keyword. And MPU21
The keyword data of the category definition database is referred to, the keywords included in the keyword data are compared with the extracted keywords, and the appearance frequency of the keywords is determined for each category. As a result, the MPU 21 sets the category in which the keyword appears most frequently as the category of the sentence. To give a simple example, the extracted keywords and the number of appearances are “weather” -5 times and “clear” -10
When the number of times is "sports" -one time, the category to which the sentence belongs is "weather forecast" and is not classified into the category of "sports".

If the appearance frequencies of all the keywords are almost the same, the category in which more keywords appear is set as the category to which the sentence belongs. If this method cannot be used to classify into categories, each category is prioritized and classified into a category with a high priority. If the text does not include any keywords included in the keyword data, the text is classified into the "other" category. When the MPU 21 determines the category of the sentence to be read out, the MPU 21 refers to the category definition database to refer to the reading setting data of the category, and performs the reading setting selection process (step S).
3). Specifically, when the classification category is “weather forecast”, the MPU 21 refers to the reading setting data and detects that the reading setting name is “weather caster”. Next, the MPU 21 reads the reading setting whose setting name data DNN of the reading setting definition database RDDB is “weather caster”. That is, MP
The U21 reads out the reading setting data such as voice type data DV = “female voice A”, speed data DS = “slightly slow”, intonation data DI = “normal”, tone data DM = “slightly polite”.

Next, the MPU 21 uses the morphological analysis dictionary DI.
With reference to C1 and the phonetic dictionary DIC2, the sentence to be read is decomposed into words and phrases, and language processing such as recognizing the connection of each word or phrase is performed (step S4).
Next, the MPU 21 refers to the waveform dictionary DIC3 and performs an acoustic process of allocating speech waveform data (speech data) to be used for reading based on the words, the clauses, and each word or the connection of the clauses that compose the sentence to be read ( Step S5). Next, the MPU 21 outputs the voice waveform data to the output unit 14. As a result, the output unit 14 performs digital / analog conversion of the voice waveform data to generate and output a read voice (step S6). Therefore, the user can listen to the voice having the voice reading characteristic suited to his / her preference.

[1.5] Modification of First Embodiment [1.5.1] First Modification In the above description, the reading setting set for each category is the reading of a preset setting. Although it could not be changed in the middle, in the first modified example, the setting can be changed even while reading aloud. Hereinafter, the first modified example will be described with reference to FIGS. 7 and 8. FIG. 7 shows a display example of the setting change dialog box. Further, FIG. 8 shows a flowchart of the reading setting changing process. When reading the text starts, MPU21
Refers to the reading setting definition database and displays the setting change dialog box 50 as shown in FIG. 7 on the display screen of the display unit during reading based on the current reading setting data (step S11). At this time, a copy of the reading setting data is generated in the RAM, and the reading process is performed based on the copied reading setting data. This is to prevent the original reading setting data in the external storage device from being changed by a temporary change.

In this case, the user can change all or part of the reading setting data by using the setting change dialog box 50. Specifically, if you want to change the speed data from the current "slow" to "slightly slow", click the select button 53 to display the selectable items in the speed data. Click to confirm your choice. On the other hand, the MPU 21 constantly monitors whether a change in the reading setting data by the user is detected (step S12). The MPU 21, when the change of the reading setting data by the user is detected (step S12; Y
es), the reading setting data in the RAM is changed (step S14).

Next, the MPU 21 reflects the setting change dialog box 50 on the category definition check box 5
It is determined whether or not 6 is checked (step S
15). When the reflection check box 56 is checked (step S15; Yes), the original reading setting data in the external storage device is changed and updated (step S16). Then, the MPU 21 determines whether or not the reading is finished (step S13), and when the reading is not finished (step S13; No),
The process proceeds to step S12 and the same process is performed. Further, when the MPU 21 does not detect the change of the reading setting data by the user (step S12; N).
o), it is determined whether or not the reading has been completed (step S
13) If the reading has not been completed (step S13; No), the process proceeds to step S12 and the same process is performed. Further, in the determination of step S13,
When the reading is finished, the setting change dialog box 50 is closed and the process is finished. If the setting name is changed by clicking the select button 51, the reading setting data corresponding to the new setting name becomes the reading setting data displayed in the setting change dialog box before the change. It does not mean that the setting data is read aloud with a different setting name. As described above, according to this modified example, the reading setting can be easily changed even during reading, which is more convenient for the user.

[1.5.2] Second Modification In the above description, a simple example in which a keyword is used for category classification has been described, but the present invention is not limited to this. For example, when selecting a part of a sentence to be read aloud, information such as the file name, the content of the sentence near the selected range, the title, the content of the image (image title, etc.), link destination information, URL, etc. It is also possible to make a comprehensive judgment to improve the accuracy of category determination.

[1.5.3] Third Modification In the above description, when the reading setting is performed, the reading setting definition database is referred to via the reading setting data of the category definition database. It is also possible to directly configure the contents of the reading setting definition database in the category definition database.

[2] Second Embodiment The second embodiment is an embodiment in which the present invention is applied to the reading of an electronic mail. In the case of electronic mail, the entire sentence is generally not long, so in the second embodiment,
The full text is read by clicking the read button.

FIG. 9 shows a display screen when an electronic mail is displayed. As shown in FIG. 9, the electronic mail display screen displayed on the display unit 13 is roughly classified into the mail operation unit 8
1, an information display section 82 and a text display section 83. The mail browser operation unit 81 is provided with various mail operation buttons 93 such as a reply button 91 for replying a mail and a transfer button 92 for transferring a mail, as well as a reading button 94 according to the present invention. The information display section 82 displays various information such as information on the sender of the currently displayed mail. The body display section 83 is an area for displaying the body of the mail to be displayed. In the second embodiment, unlike the first embodiment, since the reading range is not specified, the reading of the mail text is started by clicking the reading button 94.

In the reading setting in this case, the sender is used as the category, and the reading setting is made by the MPU 21 discriminating the sender. Therefore, if the sender is a sender who has never been set to read aloud, a dialog box inquiring the user which reading setting data to use is opened to prompt the user to make a selection. As another method of reading aloud, as in the first embodiment, it is possible to analyze the contents of the mail text, classify the categories using the characteristics of the text, and set the reading aloud. For example, as shown in FIG. 10, the category names are “casual mail”, “formal mail”, “direct mail style”, and the like. And "casual mail" is "~
right. It is selected when an expression such as "" is included. Also, "formal mail" is selected when a polite word is included. Furthermore, "direct mail style"
Is selected when a polite word is used, which starts with a greeting word as shown in FIG. The reading setting method after category classification is the same as in the first embodiment. As described above, according to the second embodiment, the read-aloud setting can be optimized according to the content of the mail and the like.

[3] Effects of the Embodiment According to the present embodiment as described above, the voice reading is set in accordance with the contents of the sentence to be read, so that the user does not have to take time and effort. Further, it is possible to comfortably use the voice reading without recognizing that it is artificial voice synthesis. In addition, since the reading aloud setting is made based on the content of the sentence, it is possible to help the user to understand the content.

[4] Modification of Embodiment [4.1] First Modification In the above description, a computer system was described as the voice reading device, but a small device incorporating a display device and a voice synthesizing device is used. The present invention can be applied to various devices such as information devices, mobile phone terminals, facsimile terminals, workstations, and wristwatch-type information devices.

[4.2] Second Modification In the above description, the case where a program for causing the computer system to function as a document reading device is stored in advance in an external storage device and expanded in RAM or the like is executed. However, it is also possible to record these programs on a removable recording medium such as an optical disc and install them, or to download and install them via a network. When downloading, it is possible to download each time it is executed.

[4.3] Third Modification In the above description, various databases such as a category definition database are stored in an external storage device, but a distributed system is configured and an external database is used. Alternatively, in the case of a small-sized information device or the like, it is possible to use an external database server.

[4.4] Fourth Modification In the above description, only the case where the same voice reading setting is used for all the reading target documents has been explained, but the conversation portion is included like novels and scenarios. In such a case, the conversation part and the other part can be set to different voice reading settings. In particular, if the speaker name and the conversational sentence are written consecutively and the speaker name and the brackets that follow it can be identified as one, as in a scenario, set the reading aloud for each speaker. It is also possible to configure. This makes it possible to read aloud with a more realistic feeling.

[4.5] Fifth Modification In the above description, the voice reading setting is made according to the contents of the document to be read, but the user is further identified and the voice reading setting is made. It is also possible to do so. For example, in a home computer, a family may be used alternately, and in such a case, the login name identifies the user. Therefore, the information for identifying whether the user is an adult or a child according to the login name is stored in advance, and even if the document has the same content, if the child is a child, read the book more slowly. If the reading is set to be the same as that given in step 1, the user can use it more comfortably.

[4.6] Sixth Modified Example In the case of a document such as haiku or tanka in which the number of characters and the reading method are defined to some extent, it is possible to make a dedicated voice reading setting. For example, in the case of haiku, voice settings may be set so as to be clearly divided into units of 5, 7, and 5.

[4.7] Seventh Modification In the above description, it is stated that the reading setting is predetermined and the setting can be changed according to the preference of the user. However, when the user changes the setting. The original settings are also saved, the state at the time of category discrimination (for example, what kind of keyword distribution was used, etc.) is learned, and if a similar discrimination state is detected, the user's change It is also possible to use the settings and use the original settings in the case of a general determination state.

[0038]

According to the present invention, even if the data for reading setting is not included in advance, the reading of voice is set according to the category to which the sentence to be read belongs. It is possible to comfortably use the voice reading without trouble and without recognizing artificial voice synthesis. Further, since the voice reading setting is made based on the category to which the sentence to be read out belongs, it is possible to help the user to understand the content.

[Brief description of drawings]

FIG. 1 is a schematic configuration block diagram of a document reading device according to an embodiment.

FIG. 2 is a schematic configuration block diagram of a document reading device main body.

FIG. 3 is a diagram for explaining an example of a browser display screen.

FIG. 4 is a diagram illustrating an example of a category definition database.

FIG. 5 is a diagram illustrating an example of a reading setting definition database.

FIG. 6 is a processing flowchart of reading processing.

FIG. 7 is a display example of a setting change dialog box.

FIG. 8 is a flowchart of a reading setting changing process.

FIG. 9 is a diagram illustrating an example of a mail display screen.

FIG. 10 is a diagram for explaining an example of another category definition database.

[Explanation of symbols]

10: Document reading device 11 ... Document reading device body 12 ... Data input section 13 ... Display section 14 ... Output section CDDB: Category definition database RDDB: Reading setting definition database

Claims (13)

[Claims] 1. A voice reading setting device for performing a voice reading of a reading target document, comprising: a discriminating unit for discriminating a category to which the reading target document belongs; and a discrimination result category for the reading target document. A voice reading setting device comprising: a setting unit that performs voice reading setting corresponding thereto. 2. The voice reading setting device according to claim 1, further comprising a setting storage unit that stores in advance voice reading settings corresponding to each category, and the setting unit refers to the setting storage unit to read the voice reading. A voice-to-speech setting device characterized by making settings. 3. The voice reading setting device according to claim 1, wherein the determination unit includes an analysis unit that analyzes the content of the reading target document. 4. The read-aloud setting device according to claim 3, wherein, when the analysis unit sets a part of the reading target document or a specific document as the reading target document, a keyword and a title included in the specific document. A voice reading setting device which analyzes the content of the reading target document based on at least one of link destination information, URL, and image content of the reading target document. 5. The voice reading setting device according to claim 1, and a voice reading unit that performs voice reading based on the reading target document data corresponding to the reading target document and the voice reading setting. A voice reading device characterized by comprising. 6. The voice reading device according to claim 5, wherein a setting display unit that displays the voice reading setting when the voice is read, a setting changing unit that allows the user to change the voice reading setting, and the voice reading. A voice reading device, comprising: a setting updating unit for reflecting a setting change on the reading setting device side. 7. A voice reading setting method for performing a voice reading of a reading target document, comprising a determination step of determining a category to which the reading target document belongs, and a determination result category for the reading target document. A voice reading setting method characterized by comprising a setting process for performing the corresponding voice reading setting. 8. The voice reading setting method according to claim 7, further comprising a setting storing step of storing in advance voice reading setting corresponding to each category, wherein the setting step stores the voice reading setting stored in the setting storing step. A voice reading setting method, wherein the voice reading setting is performed with reference to the voice reading setting. 9. The voice reading setting method according to claim 7, wherein the determining step includes an analyzing step of analyzing the content of the reading target document. 10. The voice reading setting method according to claim 9, wherein in the analyzing step, when a part of the reading target document or the specific document is set as the reading target document, a keyword and a title included in the specific document are included. A voice reading setting method characterized by analyzing the content of the reading target document based on at least one of link destination information, URL, and image content of the reading target document. 11. A voice reading setting program for setting a computer to read a voice of a reading target document, to discriminate a category to which the reading target document belongs, and to determine a category of a discrimination result for the reading target document. A voice-to-speech setting program that allows you to set the corresponding voice-to-speech settings. 12. The voice-reading setting program according to claim 11, wherein voice-reading settings corresponding to each category are stored in advance, and in the setting process, the voice-reading setting stored in the setting storing process is referred to to read the voice. A voice-to-speech setting program characterized in that it allows you to make speech settings. 13. A recording medium on which the voice reading setting program according to claim 11 or 12 is recorded.